**[OpenAI Resets GPT-5.6 Sol Usage Limits, Extends Typical Usage Duration by 18%]** According to Deep Tide TechFlow, on July 29, OpenAI reset usage quotas for all ChatGPT Work and Codex users and updated the usage limit guidelines for GPT-5.6 Sol. The company emphasized that no subscription plan quotas have been reduced. In response to previous user feedback regarding Sol consuming Codex quotas too quickly, OpenAI has implemented multiple optimizations. It is estimated that in typical usage scenarios, quota endurance will be extended by approximately 18%, with some users already noticing significant improvements starting from that day. The previously suspended 5-hour limit will also be reinstated the following day. OpenAI explained that Sol tends to engage in longer reasoning processes, initiate more tool calls, and coordinate workflows across tools and sub-agents when handling complex tasks, which may result in higher token consumption for solving high-difficulty problems. Additionally, the programmatic tool-calling mode enhances parallel processing and continuous execution during waiting periods, but it also leads to higher-than-expected single-turn response counts, cached input tokens, and overall usage. These issues are particularly evident when waiting for tool returns or conducting extensive web searches. The company stated that optimizations have been made in this regard and that efforts to improve code mode efficiency will continue. OpenAI also noted that this impact is not evenly distributed among users: the median user actually finds Sol's token efficiency to be high, but heavy users handling highly complex tasks experience significantly faster quota consumption. The company acknowledged that prior to release, it focused too much on average and median performance, neglecting long-tail high-consumption scenarios.